xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision e95faac15110125eb5d7d59e5963bc0885a41718)
1..  SPDX-License-Identifier: BSD-3-Clause
2    Copyright(c) 2010-2016 Intel Corporation.
3
4.. _testpmd_runtime:
5
6Testpmd Runtime Functions
7=========================
8
9Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10it displays a prompt that can be used to start and stop forwarding,
11configure the application, display statistics (including the extended NIC
12statistics aka xstats) , set the Flow Director and other tasks::
13
14   testpmd>
15
16The testpmd prompt has some, limited, readline support.
17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18as well as access to the command history via the up-arrow.
19
20There is also support for tab completion.
21If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
22
23.. code-block:: console
24
25   testpmd> show port <TAB>
26
27       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
31       ...
32
33
34.. note::
35
36   Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
37
38      testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39               (pause_time) (send_xon) (port_id)
40
41In the real ``testpmd>`` prompt these commands should be on a single line.
42
43Help Functions
44--------------
45
46The testpmd has on-line help for the functions that are available at runtime.
47These are divided into sections and can be accessed using help, help section or help all:
48
49.. code-block:: console
50
51   testpmd> help
52
53       help control    : Start and stop forwarding.
54       help display    : Displaying port, stats and config information.
55       help config     : Configuration information.
56       help ports      : Configuring ports.
57       help registers  : Reading and setting port registers.
58       help filters    : Filters configuration help.
59       help all        : All of the above sections.
60
61
62Command File Functions
63----------------------
64
65To facilitate loading large number of commands or to avoid cutting and pasting where not
66practical or possible testpmd supports alternative methods for executing commands.
67
68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69  will execute all CLI commands contained within the file immediately before
70  starting packet forwarding or entering interactive mode.
71
72.. code-block:: console
73
74   ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75   Interactive-mode selected
76   CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77   Configuring Port 0 (socket 0)
78   Port 0: 7C:FE:90:CB:74:CE
79   Configuring Port 1 (socket 0)
80   Port 1: 7C:FE:90:CB:74:CA
81   Checking link statuses...
82   Port 0 Link Up - speed 10000 Mbps - full-duplex
83   Port 1 Link Up - speed 10000 Mbps - full-duplex
84   Done
85   Flow rule #0 created
86   Flow rule #1 created
87   ...
88   ...
89   Flow rule #498 created
90   Flow rule #499 created
91   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
92   testpmd>
93
94
95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
96  command.
97
98.. code-block:: console
99
100   testpmd> load /home/ubuntu/flow-create-commands.txt
101   Flow rule #0 created
102   Flow rule #1 created
103   ...
104   ...
105   Flow rule #498 created
106   Flow rule #499 created
107   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
108   testpmd>
109
110
111In all cases output from any included command will be displayed as standard output.
112Execution will continue until the end of the file is reached regardless of
113whether any errors occur.  The end user must examine the output to determine if
114any failures occurred.
115
116
117Control Functions
118-----------------
119
120start
121~~~~~
122
123Start packet forwarding with current configuration::
124
125   testpmd> start
126
127start tx_first
128~~~~~~~~~~~~~~
129
130Start packet forwarding with current configuration after sending specified number of bursts of packets::
131
132   testpmd> start tx_first (""|burst_num)
133
134The default burst number is 1 when ``burst_num`` not presented.
135
136stop
137~~~~
138
139Stop packet forwarding, and display accumulated statistics::
140
141   testpmd> stop
142
143quit
144~~~~
145
146Quit to prompt::
147
148   testpmd> quit
149
150
151Display Functions
152-----------------
153
154The functions in the following sections are used to display information about the
155testpmd configuration or the NIC status.
156
157show port
158~~~~~~~~~
159
160Display information for a given port or all ports::
161
162   testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
163
164The available information categories are:
165
166* ``info``: General port information such as MAC address.
167
168* ``stats``: RX/TX statistics.
169
170* ``xstats``: RX/TX extended NIC statistics.
171
172* ``fdir``: Flow Director information and statistics.
173
174* ``stat_qmap``: Queue statistics mapping.
175
176* ``dcb_tc``: DCB information such as TC mapping.
177
178* ``cap``: Supported offload capabilities.
179
180For example:
181
182.. code-block:: console
183
184   testpmd> show port info 0
185
186   ********************* Infos for port 0 *********************
187
188   MAC address: XX:XX:XX:XX:XX:XX
189   Connect to socket: 0
190   memory allocation on the socket: 0
191   Link status: up
192   Link speed: 40000 Mbps
193   Link duplex: full-duplex
194   Promiscuous mode: enabled
195   Allmulticast mode: disabled
196   Maximum number of MAC addresses: 64
197   Maximum number of MAC addresses of hash filtering: 0
198   VLAN offload:
199       strip on
200       filter on
201       qinq(extend) off
202   Redirection table size: 512
203   Supported flow types:
204     ipv4-frag
205     ipv4-tcp
206     ipv4-udp
207     ipv4-sctp
208     ipv4-other
209     ipv6-frag
210     ipv6-tcp
211     ipv6-udp
212     ipv6-sctp
213     ipv6-other
214     l2_payload
215     port
216     vxlan
217     geneve
218     nvgre
219
220show port rss reta
221~~~~~~~~~~~~~~~~~~
222
223Display the rss redirection table entry indicated by masks on port X::
224
225   testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
226
227size is used to indicate the hardware supported reta size
228
229show port rss-hash
230~~~~~~~~~~~~~~~~~~
231
232Display the RSS hash functions and RSS hash key of a port::
233
234   testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
235
236clear port
237~~~~~~~~~~
238
239Clear the port statistics for a given port or for all ports::
240
241   testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
242
243For example::
244
245   testpmd> clear port stats all
246
247show (rxq|txq)
248~~~~~~~~~~~~~~
249
250Display information for a given port's RX/TX queue::
251
252   testpmd> show (rxq|txq) info (port_id) (queue_id)
253
254show config
255~~~~~~~~~~~
256
257Displays the configuration of the application.
258The configuration comes from the command-line, the runtime or the application defaults::
259
260   testpmd> show config (rxtx|cores|fwd|txpkts)
261
262The available information categories are:
263
264* ``rxtx``: RX/TX configuration items.
265
266* ``cores``: List of forwarding cores.
267
268* ``fwd``: Packet forwarding configuration.
269
270* ``txpkts``: Packets to TX configuration.
271
272For example:
273
274.. code-block:: console
275
276   testpmd> show config rxtx
277
278   io packet forwarding - CRC stripping disabled - packets/burst=16
279   nb forwarding cores=2 - nb forwarding ports=1
280   RX queues=1 - RX desc=128 - RX free threshold=0
281   RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282   TX queues=1 - TX desc=512 - TX free threshold=0
283   TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284   TX RS bit threshold=0 - TXQ flags=0x0
285
286set fwd
287~~~~~~~
288
289Set the packet forwarding mode::
290
291   testpmd> set fwd (io|mac|macswap|flowgen| \
292                     rxonly|txonly|csum|icmpecho) (""|retry)
293
294``retry`` can be specified for forwarding engines except ``rx_only``.
295
296The available information categories are:
297
298* ``io``: Forwards packets "as-is" in I/O mode.
299  This is the fastest possible forwarding operation as it does not access packets data.
300  This is the default mode.
301
302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303  Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304  address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305  'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
306
307* ``macswap``: MAC swap forwarding mode.
308  Swaps the source and the destination Ethernet addresses of packets before forwarding them.
309
310* ``flowgen``: Multi-flow generation mode.
311  Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
312
313* ``rxonly``: Receives packets but doesn't transmit them.
314
315* ``txonly``: Generates and transmits packets without receiving any.
316
317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
318
319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
320
321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
322
323* ``tm``: Traffic Management forwarding mode
324  Demonstrates the use of ethdev traffic management APIs and softnic PMD for
325  QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
326  available as an default option that can be enabled through CLI. The user can
327  also modify the default hierarchy or specify the new hierarchy through CLI for
328  implementing QoS scheduler.  Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
329
330Example::
331
332   testpmd> set fwd rxonly
333
334   Set rxonly packet forwarding mode
335
336
337read rxd
338~~~~~~~~
339
340Display an RX descriptor for a port RX queue::
341
342   testpmd> read rxd (port_id) (queue_id) (rxd_id)
343
344For example::
345
346   testpmd> read rxd 0 0 4
347        0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
348
349read txd
350~~~~~~~~
351
352Display a TX descriptor for a port TX queue::
353
354   testpmd> read txd (port_id) (queue_id) (txd_id)
355
356For example::
357
358   testpmd> read txd 0 0 4
359        0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
360
361ddp get list
362~~~~~~~~~~~~
363
364Get loaded dynamic device personalization (DDP) package info list::
365
366   testpmd> ddp get list (port_id)
367
368ddp get info
369~~~~~~~~~~~~
370
371Display information about dynamic device personalization (DDP) profile::
372
373   testpmd> ddp get info (profile_path)
374
375show vf stats
376~~~~~~~~~~~~~
377
378Display VF statistics::
379
380   testpmd> show vf stats (port_id) (vf_id)
381
382clear vf stats
383~~~~~~~~~~~~~~
384
385Reset VF statistics::
386
387   testpmd> clear vf stats (port_id) (vf_id)
388
389show port pctype mapping
390~~~~~~~~~~~~~~~~~~~~~~~~
391
392List all items from the pctype mapping table::
393
394   testpmd> show port (port_id) pctype mapping
395
396
397Configuration Functions
398-----------------------
399
400The testpmd application can be configured from the runtime as well as from the command-line.
401
402This section details the available configuration functions that are available.
403
404.. note::
405
406   Configuration changes only become active when forwarding is started/restarted.
407
408set default
409~~~~~~~~~~~
410
411Reset forwarding to the default configuration::
412
413   testpmd> set default
414
415set verbose
416~~~~~~~~~~~
417
418Set the debug verbosity level::
419
420   testpmd> set verbose (level)
421
422Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
423
424set log
425~~~~~~~
426
427Set the log level for a log type::
428
429	testpmd> set log global|(type) (level)
430
431Where:
432
433* ``type`` is the log name.
434
435* ``level`` is the log level.
436
437For example, to change the global log level::
438	testpmd> set log global (level)
439
440Regexes can also be used for type. To change log level of user1, user2 and user3::
441	testpmd> set log user[1-3] (level)
442
443set nbport
444~~~~~~~~~~
445
446Set the number of ports used by the application:
447
448set nbport (num)
449
450This is equivalent to the ``--nb-ports`` command-line option.
451
452set nbcore
453~~~~~~~~~~
454
455Set the number of cores used by the application::
456
457   testpmd> set nbcore (num)
458
459This is equivalent to the ``--nb-cores`` command-line option.
460
461.. note::
462
463   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
464
465set coremask
466~~~~~~~~~~~~
467
468Set the forwarding cores hexadecimal mask::
469
470   testpmd> set coremask (mask)
471
472This is equivalent to the ``--coremask`` command-line option.
473
474.. note::
475
476   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
477
478set portmask
479~~~~~~~~~~~~
480
481Set the forwarding ports hexadecimal mask::
482
483   testpmd> set portmask (mask)
484
485This is equivalent to the ``--portmask`` command-line option.
486
487set burst
488~~~~~~~~~
489
490Set number of packets per burst::
491
492   testpmd> set burst (num)
493
494This is equivalent to the ``--burst command-line`` option.
495
496When retry is enabled, the transmit delay time and number of retries can also be set::
497
498   testpmd> set burst tx delay (microseconds) retry (num)
499
500set txpkts
501~~~~~~~~~~
502
503Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
504
505   testpmd> set txpkts (x[,y]*)
506
507Where x[,y]* represents a CSV list of values, without white space.
508
509set txsplit
510~~~~~~~~~~~
511
512Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
513
514   testpmd> set txsplit (off|on|rand)
515
516Where:
517
518* ``off`` disable packet copy & split for CSUM mode.
519
520* ``on`` split outgoing packet into multiple segments. Size of each segment
521  and number of segments per packet is determined by ``set txpkts`` command
522  (see above).
523
524* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
525
526set corelist
527~~~~~~~~~~~~
528
529Set the list of forwarding cores::
530
531   testpmd> set corelist (x[,y]*)
532
533For example, to change the forwarding cores:
534
535.. code-block:: console
536
537   testpmd> set corelist 3,1
538   testpmd> show config fwd
539
540   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
541   Logical Core 3 (socket 0) forwards packets on 1 streams:
542   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
543   Logical Core 1 (socket 0) forwards packets on 1 streams:
544   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
545
546.. note::
547
548   The cores are used in the same order as specified on the command line.
549
550set portlist
551~~~~~~~~~~~~
552
553Set the list of forwarding ports::
554
555   testpmd> set portlist (x[,y]*)
556
557For example, to change the port forwarding:
558
559.. code-block:: console
560
561   testpmd> set portlist 0,2,1,3
562   testpmd> show config fwd
563
564   io packet forwarding - ports=4 - cores=1 - streams=4
565   Logical Core 3 (socket 0) forwards packets on 4 streams:
566   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
567   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
568   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
569   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
570
571set tx loopback
572~~~~~~~~~~~~~~~
573
574Enable/disable tx loopback::
575
576   testpmd> set tx loopback (port_id) (on|off)
577
578set drop enable
579~~~~~~~~~~~~~~~
580
581set drop enable bit for all queues::
582
583   testpmd> set all queues drop (port_id) (on|off)
584
585set split drop enable (for VF)
586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
587
588set split drop enable bit for VF from PF::
589
590   testpmd> set vf split drop (port_id) (vf_id) (on|off)
591
592set mac antispoof (for VF)
593~~~~~~~~~~~~~~~~~~~~~~~~~~
594
595Set mac antispoof for a VF from the PF::
596
597   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
598
599set macsec offload
600~~~~~~~~~~~~~~~~~~
601
602Enable/disable MACsec offload::
603
604   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
605   testpmd> set macsec offload (port_id) off
606
607set macsec sc
608~~~~~~~~~~~~~
609
610Configure MACsec secure connection (SC)::
611
612   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
613
614.. note::
615
616   The pi argument is ignored for tx.
617   Check the NIC Datasheet for hardware limits.
618
619set macsec sa
620~~~~~~~~~~~~~
621
622Configure MACsec secure association (SA)::
623
624   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
625
626.. note::
627
628   The IDX value must be 0 or 1.
629   Check the NIC Datasheet for hardware limits.
630
631set broadcast mode (for VF)
632~~~~~~~~~~~~~~~~~~~~~~~~~~~
633
634Set broadcast mode for a VF from the PF::
635
636   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
637
638vlan set strip
639~~~~~~~~~~~~~~
640
641Set the VLAN strip on a port::
642
643   testpmd> vlan set strip (on|off) (port_id)
644
645vlan set stripq
646~~~~~~~~~~~~~~~
647
648Set the VLAN strip for a queue on a port::
649
650   testpmd> vlan set stripq (on|off) (port_id,queue_id)
651
652vlan set stripq (for VF)
653~~~~~~~~~~~~~~~~~~~~~~~~
654
655Set VLAN strip for all queues in a pool for a VF from the PF::
656
657   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
658
659vlan set insert (for VF)
660~~~~~~~~~~~~~~~~~~~~~~~~
661
662Set VLAN insert for a VF from the PF::
663
664   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
665
666vlan set tag (for VF)
667~~~~~~~~~~~~~~~~~~~~~
668
669Set VLAN tag for a VF from the PF::
670
671   testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
672
673vlan set antispoof (for VF)
674~~~~~~~~~~~~~~~~~~~~~~~~~~~
675
676Set VLAN antispoof for a VF from the PF::
677
678   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
679
680vlan set filter
681~~~~~~~~~~~~~~~
682
683Set the VLAN filter on a port::
684
685   testpmd> vlan set filter (on|off) (port_id)
686
687vlan set qinq
688~~~~~~~~~~~~~
689
690Set the VLAN QinQ (extended queue in queue) on for a port::
691
692   testpmd> vlan set qinq (on|off) (port_id)
693
694vlan set tpid
695~~~~~~~~~~~~~
696
697Set the inner or outer VLAN TPID for packet filtering on a port::
698
699   testpmd> vlan set (inner|outer) tpid (value) (port_id)
700
701.. note::
702
703   TPID value must be a 16-bit number (value <= 65536).
704
705rx_vlan add
706~~~~~~~~~~~
707
708Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
709
710   testpmd> rx_vlan add (vlan_id|all) (port_id)
711
712.. note::
713
714   VLAN filter must be set on that port. VLAN ID < 4096.
715   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
716   in VFTA table. This is important if enabling all vlan_ids.
717
718rx_vlan rm
719~~~~~~~~~~
720
721Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
722
723   testpmd> rx_vlan rm (vlan_id|all) (port_id)
724
725rx_vlan add (for VF)
726~~~~~~~~~~~~~~~~~~~~
727
728Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
729
730   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
731
732rx_vlan rm (for VF)
733~~~~~~~~~~~~~~~~~~~
734
735Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
736
737   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
738
739tunnel_filter add
740~~~~~~~~~~~~~~~~~
741
742Add a tunnel filter on a port::
743
744   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
745            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
746            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
747
748The available information categories are:
749
750* ``vxlan``: Set tunnel type as VXLAN.
751
752* ``nvgre``: Set tunnel type as NVGRE.
753
754* ``ipingre``: Set tunnel type as IP-in-GRE.
755
756* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
757
758* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
759
760* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
761
762* ``imac``: Set filter type as Inner MAC.
763
764* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
765
766* ``oip``: Set filter type as Outer IP.
767
768* ``iip``: Set filter type as Inner IP.
769
770Example::
771
772   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
773            192.168.2.2 0 ipingre oip 1 1
774
775   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
776
777tunnel_filter remove
778~~~~~~~~~~~~~~~~~~~~
779
780Remove a tunnel filter on a port::
781
782   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
783            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
784            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
785
786rx_vxlan_port add
787~~~~~~~~~~~~~~~~~
788
789Add an UDP port for VXLAN packet filter on a port::
790
791   testpmd> rx_vxlan_port add (udp_port) (port_id)
792
793rx_vxlan_port remove
794~~~~~~~~~~~~~~~~~~~~
795
796Remove an UDP port for VXLAN packet filter on a port::
797
798   testpmd> rx_vxlan_port rm (udp_port) (port_id)
799
800tx_vlan set
801~~~~~~~~~~~
802
803Set hardware insertion of VLAN IDs in packets sent on a port::
804
805   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
806
807For example, set a single VLAN ID (5) insertion on port 0::
808
809   tx_vlan set 0 5
810
811Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
812
813   tx_vlan set 1 2 3
814
815
816tx_vlan set pvid
817~~~~~~~~~~~~~~~~
818
819Set port based hardware insertion of VLAN ID in packets sent on a port::
820
821   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
822
823tx_vlan reset
824~~~~~~~~~~~~~
825
826Disable hardware insertion of a VLAN header in packets sent on a port::
827
828   testpmd> tx_vlan reset (port_id)
829
830csum set
831~~~~~~~~
832
833Select hardware or software calculation of the checksum when
834transmitting a packet using the ``csum`` forwarding engine::
835
836   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
837
838Where:
839
840* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
841
842* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
843  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
844  supported). See also the ``csum parse-tunnel`` command.
845
846.. note::
847
848   Check the NIC Datasheet for hardware limits.
849
850RSS queue region
851~~~~~~~~~~~~~~~~
852
853Set RSS queue region span on a port::
854
855   testpmd> set port (port_id) queue-region region_id (value) \
856		queue_start_index (value) queue_num (value)
857
858Set flowtype mapping on a RSS queue region on a port::
859
860   testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
861
862where:
863
864* For the flowtype(pctype) of packet,the specific index for each type has
865  been defined in file i40e_type.h as enum i40e_filter_pctype.
866
867Set user priority mapping on a RSS queue region on a port::
868
869   testpmd> set port (port_id) queue-region UP (value) region_id (value)
870
871Flush all queue region related configuration on a port::
872
873   testpmd> set port (port_id) queue-region flush (on|off)
874
875where:
876
877* "on"is just an enable function which server for other configuration,
878  it is for all configuration about queue region from up layer,
879  at first will only keep in DPDK softwarestored in driver,
880  only after "flush on", it commit all configuration to HW.
881  "off" is just clean all configuration about queue region just now,
882  and restore all to DPDK i40e driver default config when start up.
883
884Show all queue region related configuration info on a port::
885
886   testpmd> show port (port_id) queue-region
887
888.. note::
889
890  Queue region only support on PF by now, so these command is
891  only for configuration of queue region on PF port.
892
893csum parse-tunnel
894~~~~~~~~~~~~~~~~~
895
896Define how tunneled packets should be handled by the csum forward
897engine::
898
899   testpmd> csum parse-tunnel (on|off) (tx_port_id)
900
901If enabled, the csum forward engine will try to recognize supported
902tunnel headers (vxlan, gre, ipip).
903
904If disabled, treat tunnel packets as non-tunneled packets (a inner
905header is handled as a packet payload).
906
907.. note::
908
909   The port argument is the TX port like in the ``csum set`` command.
910
911Example:
912
913Consider a packet in packet like the following::
914
915   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
916
917* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
918  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
919  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
920
921* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
922   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
923
924csum show
925~~~~~~~~~
926
927Display tx checksum offload configuration::
928
929   testpmd> csum show (port_id)
930
931tso set
932~~~~~~~
933
934Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
935
936   testpmd> tso set (segsize) (port_id)
937
938.. note::
939
940   Check the NIC datasheet for hardware limits.
941
942tso show
943~~~~~~~~
944
945Display the status of TCP Segmentation Offload::
946
947   testpmd> tso show (port_id)
948
949set port - gro
950~~~~~~~~~~~~~~
951
952Enable or disable GRO in ``csum`` forwarding engine::
953
954   testpmd> set port <port_id> gro on|off
955
956If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
957packets received from the given port.
958
959If disabled, packets received from the given port won't be performed
960GRO. By default, GRO is disabled for all ports.
961
962.. note::
963
964   When enable GRO for a port, TCP/IPv4 packets received from the port
965   will be performed GRO. After GRO, all merged packets have bad
966   checksums, since the GRO library doesn't re-calculate checksums for
967   the merged packets. Therefore, if users want the merged packets to
968   have correct checksums, please select HW IP checksum calculation and
969   HW TCP checksum calculation for the port which the merged packets are
970   transmitted to.
971
972show port - gro
973~~~~~~~~~~~~~~~
974
975Display GRO configuration for a given port::
976
977   testpmd> show port <port_id> gro
978
979set gro flush
980~~~~~~~~~~~~~
981
982Set the cycle to flush the GROed packets from reassembly tables::
983
984   testpmd> set gro flush <cycles>
985
986When enable GRO, the csum forwarding engine performs GRO on received
987packets, and the GROed packets are stored in reassembly tables. Users
988can use this command to determine when the GROed packets are flushed
989from the reassembly tables.
990
991The ``cycles`` is measured in GRO operation times. The csum forwarding
992engine flushes the GROed packets from the tables every ``cycles`` GRO
993operations.
994
995By default, the value of ``cycles`` is 1, which means flush GROed packets
996from the reassembly tables as soon as one GRO operation finishes. The value
997of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
998
999Please note that the large value of ``cycles`` may cause the poor TCP/IP
1000stack performance. Because the GROed packets are delayed to arrive the
1001stack, thus causing more duplicated ACKs and TCP retransmissions.
1002
1003set port - gso
1004~~~~~~~~~~~~~~
1005
1006Toggle per-port GSO support in ``csum`` forwarding engine::
1007
1008   testpmd> set port <port_id> gso on|off
1009
1010If enabled, the csum forwarding engine will perform GSO on supported IPv4
1011packets, transmitted on the given port.
1012
1013If disabled, packets transmitted on the given port will not undergo GSO.
1014By default, GSO is disabled for all ports.
1015
1016.. note::
1017
1018   When GSO is enabled on a port, supported IPv4 packets transmitted on that
1019   port undergo GSO. Afterwards, the segmented packets are represented by
1020   multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1021   of checksums for GSO'd segments in SW. As a result, if users want correct
1022   checksums in GSO segments, they should enable HW checksum calculation for
1023   GSO-enabled ports.
1024
1025   For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1026   by setting the following options in the csum forwarding engine:
1027
1028   testpmd> csum set outer_ip hw <port_id>
1029
1030   testpmd> csum set ip hw <port_id>
1031
1032   testpmd> csum set tcp hw <port_id>
1033
1034set gso segsz
1035~~~~~~~~~~~~~
1036
1037Set the maximum GSO segment size (measured in bytes), which includes the
1038packet header and the packet payload for GSO-enabled ports (global)::
1039
1040   testpmd> set gso segsz <length>
1041
1042show port - gso
1043~~~~~~~~~~~~~~~
1044
1045Display the status of Generic Segmentation Offload for a given port::
1046
1047   testpmd> show port <port_id> gso
1048
1049mac_addr add
1050~~~~~~~~~~~~
1051
1052Add an alternative MAC address to a port::
1053
1054   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1055
1056mac_addr remove
1057~~~~~~~~~~~~~~~
1058
1059Remove a MAC address from a port::
1060
1061   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1062
1063mac_addr add (for VF)
1064~~~~~~~~~~~~~~~~~~~~~
1065
1066Add an alternative MAC address for a VF to a port::
1067
1068   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1069
1070mac_addr set
1071~~~~~~~~~~~~
1072
1073Set the default MAC address for a port::
1074
1075   testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1076
1077mac_addr set (for VF)
1078~~~~~~~~~~~~~~~~~~~~~
1079
1080Set the MAC address for a VF from the PF::
1081
1082   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1083
1084set eth-peer
1085~~~~~~~~~~~~
1086
1087Set the forwarding peer address for certain port::
1088
1089   testpmd> set eth-peer (port_id) (perr_addr)
1090
1091This is equivalent to the ``--eth-peer`` command-line option.
1092
1093set port-uta
1094~~~~~~~~~~~~
1095
1096Set the unicast hash filter(s) on/off for a port::
1097
1098   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1099
1100set promisc
1101~~~~~~~~~~~
1102
1103Set the promiscuous mode on for a port or for all ports.
1104In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1105
1106   testpmd> set promisc (port_id|all) (on|off)
1107
1108set allmulti
1109~~~~~~~~~~~~
1110
1111Set the allmulti mode for a port or for all ports::
1112
1113   testpmd> set allmulti (port_id|all) (on|off)
1114
1115Same as the ifconfig (8) option. Controls how multicast packets are handled.
1116
1117set promisc (for VF)
1118~~~~~~~~~~~~~~~~~~~~
1119
1120Set the unicast promiscuous mode for a VF from PF.
1121It's supported by Intel i40e NICs now.
1122In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1123
1124   testpmd> set vf promisc (port_id) (vf_id) (on|off)
1125
1126set allmulticast (for VF)
1127~~~~~~~~~~~~~~~~~~~~~~~~~
1128
1129Set the multicast promiscuous mode for a VF from PF.
1130It's supported by Intel i40e NICs now.
1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1132
1133   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1134
1135set tx max bandwidth (for VF)
1136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1137
1138Set TX max absolute bandwidth (Mbps) for a VF from PF::
1139
1140   testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1141
1142set tc tx min bandwidth (for VF)
1143~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1144
1145Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1146
1147   testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1148
1149set tc tx max bandwidth (for VF)
1150~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1151
1152Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1153
1154   testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1155
1156set tc strict link priority mode
1157~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1158
1159Set some TCs' strict link priority mode on a physical port::
1160
1161   testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1162
1163set tc tx min bandwidth
1164~~~~~~~~~~~~~~~~~~~~~~~
1165
1166Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1167
1168   testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1169
1170set flow_ctrl rx
1171~~~~~~~~~~~~~~~~
1172
1173Set the link flow control parameter on a port::
1174
1175   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1176            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1177	    autoneg (on|off) (port_id)
1178
1179Where:
1180
1181* ``high_water`` (integer): High threshold value to trigger XOFF.
1182
1183* ``low_water`` (integer): Low threshold value to trigger XON.
1184
1185* ``pause_time`` (integer): Pause quota in the Pause frame.
1186
1187* ``send_xon`` (0/1): Send XON frame.
1188
1189* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1190
1191* ``autoneg``: Change the auto-negotiation parameter.
1192
1193set pfc_ctrl rx
1194~~~~~~~~~~~~~~~
1195
1196Set the priority flow control parameter on a port::
1197
1198   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1199            (pause_time) (priority) (port_id)
1200
1201Where:
1202
1203* ``high_water`` (integer): High threshold value.
1204
1205* ``low_water`` (integer): Low threshold value.
1206
1207* ``pause_time`` (integer): Pause quota in the Pause frame.
1208
1209* ``priority`` (0-7): VLAN User Priority.
1210
1211set stat_qmap
1212~~~~~~~~~~~~~
1213
1214Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1215
1216   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1217
1218For example, to set rx queue 2 on port 0 to mapping 5::
1219
1220   testpmd>set stat_qmap rx 0 2 5
1221
1222set xstats-hide-zero
1223~~~~~~~~~~~~~~~~~~~~
1224
1225Set the option to hide zero values for xstats display::
1226
1227	testpmd> set xstats-hide-zero on|off
1228
1229.. note::
1230
1231	By default, the zero values are displayed for xstats.
1232
1233set port - rx/tx (for VF)
1234~~~~~~~~~~~~~~~~~~~~~~~~~
1235
1236Set VF receive/transmit from a port::
1237
1238   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1239
1240set port - mac address filter (for VF)
1241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1242
1243Add/Remove unicast or multicast MAC addr filter for a VF::
1244
1245   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1246            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1247
1248set port - rx mode(for VF)
1249~~~~~~~~~~~~~~~~~~~~~~~~~~
1250
1251Set the VF receive mode of a port::
1252
1253   testpmd> set port (port_id) vf (vf_id) \
1254            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1255
1256The available receive modes are:
1257
1258* ``AUPE``: Accepts untagged VLAN.
1259
1260* ``ROPE``: Accepts unicast hash.
1261
1262* ``BAM``: Accepts broadcast packets.
1263
1264* ``MPE``: Accepts all multicast packets.
1265
1266set port - tx_rate (for Queue)
1267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1268
1269Set TX rate limitation for a queue on a port::
1270
1271   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1272
1273set port - tx_rate (for VF)
1274~~~~~~~~~~~~~~~~~~~~~~~~~~~
1275
1276Set TX rate limitation for queues in VF on a port::
1277
1278   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1279
1280set port - mirror rule
1281~~~~~~~~~~~~~~~~~~~~~~
1282
1283Set pool or vlan type mirror rule for a port::
1284
1285   testpmd> set port (port_id) mirror-rule (rule_id) \
1286            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1287            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1288
1289Set link mirror rule for a port::
1290
1291   testpmd> set port (port_id) mirror-rule (rule_id) \
1292           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1293
1294For example to enable mirror traffic with vlan 0,1 to pool 0::
1295
1296   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1297
1298reset port - mirror rule
1299~~~~~~~~~~~~~~~~~~~~~~~~
1300
1301Reset a mirror rule for a port::
1302
1303   testpmd> reset port (port_id) mirror-rule (rule_id)
1304
1305set flush_rx
1306~~~~~~~~~~~~
1307
1308Set the flush on RX streams before forwarding.
1309The default is flush ``on``.
1310Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1311
1312   testpmd> set flush_rx off
1313
1314set bypass mode
1315~~~~~~~~~~~~~~~
1316
1317Set the bypass mode for the lowest port on bypass enabled NIC::
1318
1319   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1320
1321set bypass event
1322~~~~~~~~~~~~~~~~
1323
1324Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1325
1326   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1327            mode (normal|bypass|isolate) (port_id)
1328
1329Where:
1330
1331* ``timeout``: Enable bypass after watchdog timeout.
1332
1333* ``os_on``: Enable bypass when OS/board is powered on.
1334
1335* ``os_off``: Enable bypass when OS/board is powered off.
1336
1337* ``power_on``: Enable bypass when power supply is turned on.
1338
1339* ``power_off``: Enable bypass when power supply is turned off.
1340
1341
1342set bypass timeout
1343~~~~~~~~~~~~~~~~~~
1344
1345Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1346
1347   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1348
1349show bypass config
1350~~~~~~~~~~~~~~~~~~
1351
1352Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1353
1354   testpmd> show bypass config (port_id)
1355
1356set link up
1357~~~~~~~~~~~
1358
1359Set link up for a port::
1360
1361   testpmd> set link-up port (port id)
1362
1363set link down
1364~~~~~~~~~~~~~
1365
1366Set link down for a port::
1367
1368   testpmd> set link-down port (port id)
1369
1370E-tag set
1371~~~~~~~~~
1372
1373Enable E-tag insertion for a VF on a port::
1374
1375   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1376
1377Disable E-tag insertion for a VF on a port::
1378
1379   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1380
1381Enable/disable E-tag stripping on a port::
1382
1383   testpmd> E-tag set stripping (on|off) port (port_id)
1384
1385Enable/disable E-tag based forwarding on a port::
1386
1387   testpmd> E-tag set forwarding (on|off) port (port_id)
1388
1389Add an E-tag forwarding filter on a port::
1390
1391   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1392
1393Delete an E-tag forwarding filter on a port::
1394   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1395
1396ddp add
1397~~~~~~~
1398
1399Load a dynamic device personalization (DDP) profile and store backup profile::
1400
1401   testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1402
1403ddp del
1404~~~~~~~
1405
1406Delete a dynamic device personalization profile and restore backup profile::
1407
1408   testpmd> ddp del (port_id) (backup_profile_path)
1409
1410ptype mapping
1411~~~~~~~~~~~~~
1412
1413List all items from the ptype mapping table::
1414
1415   testpmd> ptype mapping get (port_id) (valid_only)
1416
1417Where:
1418
1419* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1420
1421Replace a specific or a group of software defined ptype with a new one::
1422
1423   testpmd> ptype mapping replace  (port_id) (target) (mask) (pkt_type)
1424
1425where:
1426
1427* ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1428
1429* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1430
1431* ``pkt_type``: The new software ptype to replace the old ones.
1432
1433Update hardware defined ptype to software defined packet type mapping table::
1434
1435   testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1436
1437where:
1438
1439* ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1440
1441* ``sw_ptype``: software ptype as the value of the ptype mapping table.
1442
1443Reset ptype mapping table::
1444
1445   testpmd> ptype mapping reset (port_id)
1446
1447Port Functions
1448--------------
1449
1450The following sections show functions for configuring ports.
1451
1452.. note::
1453
1454   Port configuration changes only become active when forwarding is started/restarted.
1455
1456port attach
1457~~~~~~~~~~~
1458
1459Attach a port specified by pci address or virtual device args::
1460
1461   testpmd> port attach (identifier)
1462
1463To attach a new pci device, the device should be recognized by kernel first.
1464Then it should be moved under DPDK management.
1465Finally the port can be attached to testpmd.
1466
1467For example, to move a pci device using ixgbe under DPDK management:
1468
1469.. code-block:: console
1470
1471   # Check the status of the available devices.
1472   ./usertools/dpdk-devbind.py --status
1473
1474   Network devices using DPDK-compatible driver
1475   ============================================
1476   <none>
1477
1478   Network devices using kernel driver
1479   ===================================
1480   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1481
1482
1483   # Bind the device to igb_uio.
1484   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1485
1486
1487   # Recheck the status of the devices.
1488   ./usertools/dpdk-devbind.py --status
1489   Network devices using DPDK-compatible driver
1490   ============================================
1491   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1492
1493To attach a port created by virtual device, above steps are not needed.
1494
1495For example, to attach a port whose pci address is 0000:0a:00.0.
1496
1497.. code-block:: console
1498
1499   testpmd> port attach 0000:0a:00.0
1500   Attaching a new port...
1501   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1502   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1503   EAL:   PCI memory mapped at 0x7f83bfa00000
1504   EAL:   PCI memory mapped at 0x7f83bfa80000
1505   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1506   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1507   Port 0 is attached. Now total ports is 1
1508   Done
1509
1510For example, to attach a port created by pcap PMD.
1511
1512.. code-block:: console
1513
1514   testpmd> port attach net_pcap0
1515   Attaching a new port...
1516   PMD: Initializing pmd_pcap for net_pcap0
1517   PMD: Creating pcap-backed ethdev on numa socket 0
1518   Port 0 is attached. Now total ports is 1
1519   Done
1520
1521In this case, identifier is ``net_pcap0``.
1522This identifier format is the same as ``--vdev`` format of DPDK applications.
1523
1524For example, to re-attach a bonded port which has been previously detached,
1525the mode and slave parameters must be given.
1526
1527.. code-block:: console
1528
1529   testpmd> port attach net_bond_0,mode=0,slave=1
1530   Attaching a new port...
1531   EAL: Initializing pmd_bond for net_bond_0
1532   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1533   Port 0 is attached. Now total ports is 1
1534   Done
1535
1536
1537port detach
1538~~~~~~~~~~~
1539
1540Detach a specific port::
1541
1542   testpmd> port detach (port_id)
1543
1544Before detaching a port, the port should be stopped and closed.
1545
1546For example, to detach a pci device port 0.
1547
1548.. code-block:: console
1549
1550   testpmd> port stop 0
1551   Stopping ports...
1552   Done
1553   testpmd> port close 0
1554   Closing ports...
1555   Done
1556
1557   testpmd> port detach 0
1558   Detaching a port...
1559   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1560   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1561   EAL:   PCI memory unmapped at 0x7f83bfa00000
1562   EAL:   PCI memory unmapped at 0x7f83bfa80000
1563   Done
1564
1565
1566For example, to detach a virtual device port 0.
1567
1568.. code-block:: console
1569
1570   testpmd> port stop 0
1571   Stopping ports...
1572   Done
1573   testpmd> port close 0
1574   Closing ports...
1575   Done
1576
1577   testpmd> port detach 0
1578   Detaching a port...
1579   PMD: Closing pcap ethdev on numa socket 0
1580   Port 'net_pcap0' is detached. Now total ports is 0
1581   Done
1582
1583To remove a pci device completely from the system, first detach the port from testpmd.
1584Then the device should be moved under kernel management.
1585Finally the device can be removed using kernel pci hotplug functionality.
1586
1587For example, to move a pci device under kernel management:
1588
1589.. code-block:: console
1590
1591   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1592
1593   ./usertools/dpdk-devbind.py --status
1594
1595   Network devices using DPDK-compatible driver
1596   ============================================
1597   <none>
1598
1599   Network devices using kernel driver
1600   ===================================
1601   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1602
1603To remove a port created by a virtual device, above steps are not needed.
1604
1605port start
1606~~~~~~~~~~
1607
1608Start all ports or a specific port::
1609
1610   testpmd> port start (port_id|all)
1611
1612port stop
1613~~~~~~~~~
1614
1615Stop all ports or a specific port::
1616
1617   testpmd> port stop (port_id|all)
1618
1619port close
1620~~~~~~~~~~
1621
1622Close all ports or a specific port::
1623
1624   testpmd> port close (port_id|all)
1625
1626port config - queue ring size
1627~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1628
1629Configure a rx/tx queue ring size::
1630
1631   testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1632
1633Only take effect after command that (re-)start the port or command that setup specific queue.
1634
1635port start/stop queue
1636~~~~~~~~~~~~~~~~~~~~~
1637
1638Start/stop a rx/tx queue on a specific port::
1639
1640   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1641
1642port setup queue
1643~~~~~~~~~~~~~~~~~~~~~
1644
1645Setup a rx/tx queue on a specific port::
1646
1647   testpmd> port (port_id) (rxq|txq) (queue_id) setup
1648
1649Only take effect when port is started.
1650
1651port config - speed
1652~~~~~~~~~~~~~~~~~~~
1653
1654Set the speed and duplex mode for all ports or a specific port::
1655
1656   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1657            duplex (half|full|auto)
1658
1659port config - queues/descriptors
1660~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1661
1662Set number of queues/descriptors for rxq, txq, rxd and txd::
1663
1664   testpmd> port config all (rxq|txq|rxd|txd) (value)
1665
1666This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1667
1668port config - max-pkt-len
1669~~~~~~~~~~~~~~~~~~~~~~~~~
1670
1671Set the maximum packet length::
1672
1673   testpmd> port config all max-pkt-len (value)
1674
1675This is equivalent to the ``--max-pkt-len`` command-line option.
1676
1677port config - CRC Strip
1678~~~~~~~~~~~~~~~~~~~~~~~
1679
1680Set hardware CRC stripping on or off for all ports::
1681
1682   testpmd> port config all crc-strip (on|off)
1683
1684CRC stripping is on by default.
1685
1686The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1687
1688port config - scatter
1689~~~~~~~~~~~~~~~~~~~~~~~
1690
1691Set RX scatter mode on or off for all ports::
1692
1693   testpmd> port config all scatter (on|off)
1694
1695RX scatter mode is off by default.
1696
1697The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1698
1699port config - RX Checksum
1700~~~~~~~~~~~~~~~~~~~~~~~~~
1701
1702Set hardware RX checksum offload to on or off for all ports::
1703
1704   testpmd> port config all rx-cksum (on|off)
1705
1706Checksum offload is off by default.
1707
1708The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1709
1710port config - VLAN
1711~~~~~~~~~~~~~~~~~~
1712
1713Set hardware VLAN on or off for all ports::
1714
1715   testpmd> port config all hw-vlan (on|off)
1716
1717Hardware VLAN is off by default.
1718
1719The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1720
1721port config - VLAN filter
1722~~~~~~~~~~~~~~~~~~~~~~~~~
1723
1724Set hardware VLAN filter on or off for all ports::
1725
1726   testpmd> port config all hw-vlan-filter (on|off)
1727
1728Hardware VLAN filter is off by default.
1729
1730The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1731
1732port config - VLAN strip
1733~~~~~~~~~~~~~~~~~~~~~~~~
1734
1735Set hardware VLAN strip on or off for all ports::
1736
1737   testpmd> port config all hw-vlan-strip (on|off)
1738
1739Hardware VLAN strip is off by default.
1740
1741The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1742
1743port config - VLAN extend
1744~~~~~~~~~~~~~~~~~~~~~~~~~
1745
1746Set hardware VLAN extend on or off for all ports::
1747
1748   testpmd> port config all hw-vlan-extend (on|off)
1749
1750Hardware VLAN extend is off by default.
1751
1752The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1753
1754port config - Drop Packets
1755~~~~~~~~~~~~~~~~~~~~~~~~~~
1756
1757Set packet drop for packets with no descriptors on or off for all ports::
1758
1759   testpmd> port config all drop-en (on|off)
1760
1761Packet dropping for packets with no descriptors is off by default.
1762
1763The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1764
1765port config - RSS
1766~~~~~~~~~~~~~~~~~
1767
1768Set the RSS (Receive Side Scaling) mode on or off::
1769
1770   testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1771
1772RSS is on by default.
1773
1774The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1775The ``default`` option enables all supported RSS types reported by device info.
1776The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1777
1778port config - RSS Reta
1779~~~~~~~~~~~~~~~~~~~~~~
1780
1781Set the RSS (Receive Side Scaling) redirection table::
1782
1783   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1784
1785port config - DCB
1786~~~~~~~~~~~~~~~~~
1787
1788Set the DCB mode for an individual port::
1789
1790   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1791
1792The traffic class should be 4 or 8.
1793
1794port config - Burst
1795~~~~~~~~~~~~~~~~~~~
1796
1797Set the number of packets per burst::
1798
1799   testpmd> port config all burst (value)
1800
1801This is equivalent to the ``--burst`` command-line option.
1802
1803port config - Threshold
1804~~~~~~~~~~~~~~~~~~~~~~~
1805
1806Set thresholds for TX/RX queues::
1807
1808   testpmd> port config all (threshold) (value)
1809
1810Where the threshold type can be:
1811
1812* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1813
1814* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1815
1816* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1817
1818* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1819
1820* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1821
1822* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1823
1824* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1825
1826* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1827
1828* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1829
1830These threshold options are also available from the command-line.
1831
1832port config - E-tag
1833~~~~~~~~~~~~~~~~~~~
1834
1835Set the value of ether-type for E-tag::
1836
1837   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1838
1839Enable/disable the E-tag support::
1840
1841   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1842
1843port config pctype mapping
1844~~~~~~~~~~~~~~~~~~~~~~~~~~
1845
1846Reset pctype mapping table::
1847
1848   testpmd> port config (port_id) pctype mapping reset
1849
1850Update hardware defined pctype to software defined flow type mapping table::
1851
1852   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1853
1854where:
1855
1856* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1857
1858* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1859
1860port config input set
1861~~~~~~~~~~~~~~~~~~~~~
1862
1863Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1864   testpmd> port config (port_id) pctype (pctype_id) \
1865            (hash_inset|fdir_inset|fdir_flx_inset) \
1866	    (get|set|clear) field (field_idx)
1867
1868Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1869   testpmd> port config (port_id) pctype (pctype_id) \
1870            (hash_inset|fdir_inset|fdir_flx_inset) clear all
1871
1872where:
1873
1874* ``pctype_id``: hardware packet classification types.
1875* ``field_idx``: hardware field index.
1876
1877port config udp_tunnel_port
1878~~~~~~~~~~~~~~~~~~~~~~~~~~~
1879
1880Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
1881    testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
1882
1883Link Bonding Functions
1884----------------------
1885
1886The Link Bonding functions make it possible to dynamically create and
1887manage link bonding devices from within testpmd interactive prompt.
1888
1889create bonded device
1890~~~~~~~~~~~~~~~~~~~~
1891
1892Create a new bonding device::
1893
1894   testpmd> create bonded device (mode) (socket)
1895
1896For example, to create a bonded device in mode 1 on socket 0::
1897
1898   testpmd> create bonded 1 0
1899   created new bonded device (port X)
1900
1901add bonding slave
1902~~~~~~~~~~~~~~~~~
1903
1904Adds Ethernet device to a Link Bonding device::
1905
1906   testpmd> add bonding slave (slave id) (port id)
1907
1908For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1909
1910   testpmd> add bonding slave 6 10
1911
1912
1913remove bonding slave
1914~~~~~~~~~~~~~~~~~~~~
1915
1916Removes an Ethernet slave device from a Link Bonding device::
1917
1918   testpmd> remove bonding slave (slave id) (port id)
1919
1920For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1921
1922   testpmd> remove bonding slave 6 10
1923
1924set bonding mode
1925~~~~~~~~~~~~~~~~
1926
1927Set the Link Bonding mode of a Link Bonding device::
1928
1929   testpmd> set bonding mode (value) (port id)
1930
1931For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1932
1933   testpmd> set bonding mode 3 10
1934
1935set bonding primary
1936~~~~~~~~~~~~~~~~~~~
1937
1938Set an Ethernet slave device as the primary device on a Link Bonding device::
1939
1940   testpmd> set bonding primary (slave id) (port id)
1941
1942For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1943
1944   testpmd> set bonding primary 6 10
1945
1946set bonding mac
1947~~~~~~~~~~~~~~~
1948
1949Set the MAC address of a Link Bonding device::
1950
1951   testpmd> set bonding mac (port id) (mac)
1952
1953For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1954
1955   testpmd> set bonding mac 10 00:00:00:00:00:01
1956
1957set bonding xmit_balance_policy
1958~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1959
1960Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1961
1962   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1963
1964For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1965
1966   testpmd> set bonding xmit_balance_policy 10 l34
1967
1968
1969set bonding mon_period
1970~~~~~~~~~~~~~~~~~~~~~~
1971
1972Set the link status monitoring polling period in milliseconds for a bonding device.
1973
1974This adds support for PMD slave devices which do not support link status interrupts.
1975When the mon_period is set to a value greater than 0 then all PMD's which do not support
1976link status ISR will be queried every polling interval to check if their link status has changed::
1977
1978   testpmd> set bonding mon_period (port_id) (value)
1979
1980For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1981
1982   testpmd> set bonding mon_period 5 150
1983
1984
1985set bonding lacp dedicated_queue
1986~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1987
1988Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1989when in mode 4 (link-aggregration-802.3ad)::
1990
1991   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1992
1993
1994set bonding agg_mode
1995~~~~~~~~~~~~~~~~~~~~
1996
1997Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1998
1999   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2000
2001
2002show bonding config
2003~~~~~~~~~~~~~~~~~~~
2004
2005Show the current configuration of a Link Bonding device::
2006
2007   testpmd> show bonding config (port id)
2008
2009For example,
2010to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2011in balance mode with a transmission policy of layer 2+3::
2012
2013   testpmd> show bonding config 9
2014        Bonding mode: 2
2015        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2016        Slaves (3): [1 3 4]
2017        Active Slaves (3): [1 3 4]
2018        Primary: [3]
2019
2020
2021Register Functions
2022------------------
2023
2024The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2025This is mainly useful for debugging purposes.
2026Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2027and fields that can be accessed.
2028
2029read reg
2030~~~~~~~~
2031
2032Display the value of a port register::
2033
2034   testpmd> read reg (port_id) (address)
2035
2036For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2037
2038   testpmd> read reg 0 0xEE00
2039   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2040
2041read regfield
2042~~~~~~~~~~~~~
2043
2044Display a port register bit field::
2045
2046   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2047
2048For example, reading the lowest two bits from the register in the example above::
2049
2050   testpmd> read regfield 0 0xEE00 0 1
2051   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2052
2053read regbit
2054~~~~~~~~~~~
2055
2056Display a single port register bit::
2057
2058   testpmd> read regbit (port_id) (address) (bit_x)
2059
2060For example, reading the lowest bit from the register in the example above::
2061
2062   testpmd> read regbit 0 0xEE00 0
2063   port 0 PCI register at offset 0xEE00: bit 0=1
2064
2065write reg
2066~~~~~~~~~
2067
2068Set the value of a port register::
2069
2070   testpmd> write reg (port_id) (address) (value)
2071
2072For example, to clear a register::
2073
2074   testpmd> write reg 0 0xEE00 0x0
2075   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2076
2077write regfield
2078~~~~~~~~~~~~~~
2079
2080Set bit field of a port register::
2081
2082   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2083
2084For example, writing to the register cleared in the example above::
2085
2086   testpmd> write regfield 0 0xEE00 0 1 2
2087   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2088
2089write regbit
2090~~~~~~~~~~~~
2091
2092Set single bit value of a port register::
2093
2094   testpmd> write regbit (port_id) (address) (bit_x) (value)
2095
2096For example, to set the high bit in the register from the example above::
2097
2098   testpmd> write regbit 0 0xEE00 31 1
2099   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2100
2101Traffic Metering and Policing
2102-----------------------------
2103
2104The following section shows functions for configuring traffic metering and
2105policing on the ethernet device through the use of generic ethdev API.
2106
2107show port traffic management capability
2108~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2109
2110Show traffic metering and policing capability of the port::
2111
2112   testpmd> show port meter cap (port_id)
2113
2114add port meter profile (srTCM rfc2967)
2115~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2116
2117Add meter profile (srTCM rfc2697) to the ethernet device::
2118
2119   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2120   (cir) (cbs) (ebs)
2121
2122where:
2123
2124* ``profile_id``: ID for the meter profile.
2125* ``cir``: Committed Information Rate (CIR) (bytes/second).
2126* ``cbs``: Committed Burst Size (CBS) (bytes).
2127* ``ebs``: Excess Burst Size (EBS) (bytes).
2128
2129add port meter profile (trTCM rfc2968)
2130~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2131
2132Add meter profile (srTCM rfc2698) to the ethernet device::
2133
2134   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2135   (cir) (pir) (cbs) (pbs)
2136
2137where:
2138
2139* ``profile_id``: ID for the meter profile.
2140* ``cir``: Committed information rate (bytes/second).
2141* ``pir``: Peak information rate (bytes/second).
2142* ``cbs``: Committed burst size (bytes).
2143* ``pbs``: Peak burst size (bytes).
2144
2145add port meter profile (trTCM rfc4115)
2146~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2147
2148Add meter profile (trTCM rfc4115) to the ethernet device::
2149
2150   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2151   (cir) (eir) (cbs) (ebs)
2152
2153where:
2154
2155* ``profile_id``: ID for the meter profile.
2156* ``cir``: Committed information rate (bytes/second).
2157* ``eir``: Excess information rate (bytes/second).
2158* ``cbs``: Committed burst size (bytes).
2159* ``ebs``: Excess burst size (bytes).
2160
2161delete port meter profile
2162~~~~~~~~~~~~~~~~~~~~~~~~~
2163
2164Delete meter profile from the ethernet device::
2165
2166   testpmd> del port meter profile (port_id) (profile_id)
2167
2168create port meter
2169~~~~~~~~~~~~~~~~~
2170
2171Create new meter object for the ethernet device::
2172
2173   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2174   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2175   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2176   (dscp_tbl_entry63)]
2177
2178where:
2179
2180* ``mtr_id``: meter object ID.
2181* ``profile_id``: ID for the meter profile.
2182* ``meter_enable``: When this parameter has a non-zero value, the meter object
2183  gets enabled at the time of creation, otherwise remains disabled.
2184* ``g_action``: Policer action for the packet with green color.
2185* ``y_action``: Policer action for the packet with yellow color.
2186* ``r_action``: Policer action for the packet with red color.
2187* ``stats_mask``: Mask of statistics counter types to be enabled for the
2188  meter object.
2189* ``shared``:  When this parameter has a non-zero value, the meter object is
2190  shared by multiple flows. Otherwise, meter object is used by single flow.
2191* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2192  input color for the current meter object is determined by the latest meter
2193  object in the same flow. Otherwise, the current meter object uses the
2194  *dscp_table* to determine the input color.
2195* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2196  color, 0 <= x <= 63.
2197
2198enable port meter
2199~~~~~~~~~~~~~~~~~
2200
2201Enable meter for the ethernet device::
2202
2203   testpmd> enable port meter (port_id) (mtr_id)
2204
2205disable port meter
2206~~~~~~~~~~~~~~~~~~
2207
2208Disable meter for the ethernet device::
2209
2210   testpmd> disable port meter (port_id) (mtr_id)
2211
2212delete port meter
2213~~~~~~~~~~~~~~~~~
2214
2215Delete meter for the ethernet device::
2216
2217   testpmd> del port meter (port_id) (mtr_id)
2218
2219Set port meter profile
2220~~~~~~~~~~~~~~~~~~~~~~
2221
2222Set meter profile for the ethernet device::
2223
2224   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2225
2226set port meter dscp table
2227~~~~~~~~~~~~~~~~~~~~~~~~~
2228
2229Set meter dscp table for the ethernet device::
2230
2231   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2232   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2233
2234set port meter policer action
2235~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2236
2237Set meter policer action for the ethernet device::
2238
2239   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2240   (action0) [(action1) (action1)]
2241
2242where:
2243
2244* ``action_mask``: Bit mask indicating which policer actions need to be
2245  updated. One or more policer actions can be updated in a single function
2246  invocation. To update the policer action associated with color C, bit
2247  (1 << C) needs to be set in *action_mask* and element at position C
2248  in the *actions* array needs to be valid.
2249* ``actionx``: Policer action for the color x,
2250  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2251
2252set port meter stats mask
2253~~~~~~~~~~~~~~~~~~~~~~~~~
2254
2255Set meter stats mask for the ethernet device::
2256
2257   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2258
2259where:
2260
2261* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2262
2263show port meter stats
2264~~~~~~~~~~~~~~~~~~~~~
2265
2266Show meter stats of the ethernet device::
2267
2268   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2269
2270where:
2271
2272* ``clear``: Flag that indicates whether the statistics counters should
2273  be cleared (i.e. set to zero) immediately after they have been read or not.
2274
2275Traffic Management
2276------------------
2277
2278The following section shows functions for configuring traffic management on
2279on the ethernet device through the use of generic TM API.
2280
2281show port traffic management capability
2282~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2283
2284Show traffic management capability of the port::
2285
2286   testpmd> show port tm cap (port_id)
2287
2288show port traffic management capability (hierarchy level)
2289~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2290
2291Show traffic management hierarchy level capability of the port::
2292
2293   testpmd> show port tm level cap (port_id) (level_id)
2294
2295show port traffic management capability (hierarchy node level)
2296~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2297
2298Show the traffic management hierarchy node capability of the port::
2299
2300   testpmd> show port tm node cap (port_id) (node_id)
2301
2302show port traffic management hierarchy node type
2303~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2304
2305Show the port traffic management hierarchy node type::
2306
2307   testpmd> show port tm node type (port_id) (node_id)
2308
2309show port traffic management hierarchy node stats
2310~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2311
2312Show the port traffic management hierarchy node statistics::
2313
2314   testpmd> show port tm node stats (port_id) (node_id) (clear)
2315
2316where:
2317
2318* ``clear``: When this parameter has a non-zero value, the statistics counters
2319  are cleared (i.e. set to zero) immediately after they have been read,
2320  otherwise the statistics counters are left untouched.
2321
2322Add port traffic management private shaper profile
2323~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2324
2325Add the port traffic management private shaper profile::
2326
2327   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2328   (tb_rate) (tb_size) (packet_length_adjust)
2329
2330where:
2331
2332* ``shaper_profile id``: Shaper profile ID for the new profile.
2333* ``tb_rate``: Token bucket rate (bytes per second).
2334* ``tb_size``: Token bucket size (bytes).
2335* ``packet_length_adjust``: The value (bytes) to be added to the length of
2336  each packet for the purpose of shaping. This parameter value can be used to
2337  correct the packet length with the framing overhead bytes that are consumed
2338  on the wire.
2339
2340Delete port traffic management private shaper profile
2341~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2342
2343Delete the port traffic management private shaper::
2344
2345   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2346
2347where:
2348
2349* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2350
2351Add port traffic management shared shaper
2352~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2353
2354Create the port traffic management shared shaper::
2355
2356   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2357   (shaper_profile_id)
2358
2359where:
2360
2361* ``shared_shaper_id``: Shared shaper ID to be created.
2362* ``shaper_profile id``: Shaper profile ID for shared shaper.
2363
2364Set port traffic management shared shaper
2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2366
2367Update the port traffic management shared shaper::
2368
2369   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2370   (shaper_profile_id)
2371
2372where:
2373
2374* ``shared_shaper_id``: Shared shaper ID to be update.
2375* ``shaper_profile id``: Shaper profile ID for shared shaper.
2376
2377Delete port traffic management shared shaper
2378~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2379
2380Delete the port traffic management shared shaper::
2381
2382   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2383
2384where:
2385
2386* ``shared_shaper_id``: Shared shaper ID to be deleted.
2387
2388Set port traffic management hiearchy node private shaper
2389~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2390
2391set the port traffic management hierarchy node private shaper::
2392
2393   testpmd> set port tm node shaper profile (port_id) (node_id) \
2394   (shaper_profile_id)
2395
2396where:
2397
2398* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2399  hierarchy node.
2400
2401Add port traffic management WRED profile
2402~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2403
2404Create a new WRED profile::
2405
2406   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2407   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2408   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2409   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2410
2411where:
2412
2413* ``wred_profile id``: Identifier for the newly create WRED profile
2414* ``color_g``: Packet color (green)
2415* ``min_th_g``: Minimum queue threshold for packet with green color
2416* ``max_th_g``: Minimum queue threshold for packet with green color
2417* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2418* ``wq_log2_g``: Negated log2 of queue weight (wq)
2419* ``color_y``: Packet color (yellow)
2420* ``min_th_y``: Minimum queue threshold for packet with yellow color
2421* ``max_th_y``: Minimum queue threshold for packet with yellow color
2422* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2423* ``wq_log2_y``: Negated log2 of queue weight (wq)
2424* ``color_r``: Packet color (red)
2425* ``min_th_r``: Minimum queue threshold for packet with yellow color
2426* ``max_th_r``: Minimum queue threshold for packet with yellow color
2427* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2428* ``wq_log2_r``: Negated log2 of queue weight (wq)
2429
2430Delete port traffic management WRED profile
2431~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2432
2433Delete the WRED profile::
2434
2435   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2436
2437Add port traffic management hierarchy nonleaf node
2438~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2439
2440Add nonleaf node to port traffic management hiearchy::
2441
2442   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2443   (priority) (weight) (level_id) (shaper_profile_id) \
2444   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2445   [(shared_shaper_0) (shared_shaper_1) ...] \
2446
2447where:
2448
2449* ``parent_node_id``: Node ID of the parent.
2450* ``priority``: Node priority (highest node priority is zero). This is used by
2451  the SP algorithm running on the parent node for scheduling this node.
2452* ``weight``: Node weight (lowest weight is one). The node weight is relative
2453  to the weight sum of all siblings that have the same priority. It is used by
2454  the WFQ algorithm running on the parent node for scheduling this node.
2455* ``level_id``: Hiearchy level of the node.
2456* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2457  the node.
2458* ``n_sp_priorities``: Number of strict priorities.
2459* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2460* ``n_shared_shapers``: Number of shared shapers.
2461* ``shared_shaper_id``: Shared shaper id.
2462
2463Add port traffic management hierarchy leaf node
2464~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2465
2466Add leaf node to port traffic management hiearchy::
2467
2468   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2469   (priority) (weight) (level_id) (shaper_profile_id) \
2470   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2471   [(shared_shaper_id) (shared_shaper_id) ...] \
2472
2473where:
2474
2475* ``parent_node_id``: Node ID of the parent.
2476* ``priority``: Node priority (highest node priority is zero). This is used by
2477  the SP algorithm running on the parent node for scheduling this node.
2478* ``weight``: Node weight (lowest weight is one). The node weight is relative
2479  to the weight sum of all siblings that have the same priority. It is used by
2480  the WFQ algorithm running on the parent node for scheduling this node.
2481* ``level_id``: Hiearchy level of the node.
2482* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2483  the node.
2484* ``cman_mode``: Congestion management mode to be enabled for this node.
2485* ``wred_profile_id``: WRED profile id to be enabled for this node.
2486* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2487* ``n_shared_shapers``: Number of shared shapers.
2488* ``shared_shaper_id``: Shared shaper id.
2489
2490Delete port traffic management hierarchy node
2491~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2492
2493Delete node from port traffic management hiearchy::
2494
2495   testpmd> del port tm node (port_id) (node_id)
2496
2497Update port traffic management hierarchy parent node
2498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2499
2500Update port traffic management hierarchy parent node::
2501
2502   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2503   (priority) (weight)
2504
2505This function can only be called after the hierarchy commit invocation. Its
2506success depends on the port support for this operation, as advertised through
2507the port capability set. This function is valid for all nodes of the traffic
2508management hierarchy except root node.
2509
2510Suspend port traffic management hierarchy node
2511~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2512
2513   testpmd> suspend port tm node (port_id) (node_id)
2514
2515Resume port traffic management hierarchy node
2516~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2517
2518   testpmd> resume port tm node (port_id) (node_id)
2519
2520Commit port traffic management hierarchy
2521~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2522
2523Commit the traffic management hierarchy on the port::
2524
2525   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2526
2527where:
2528
2529* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2530  call failure. On the other hand, hierarchy is preserved when this parameter
2531  is equal to zero.
2532
2533Set port traffic management default hierarchy (tm forwarding mode)
2534~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2535
2536set the traffic management default hierarchy on the port::
2537
2538   testpmd> set port tm hierarchy default (port_id)
2539
2540Filter Functions
2541----------------
2542
2543This section details the available filter functions that are available.
2544
2545Note these functions interface the deprecated legacy filtering framework,
2546superseded by *rte_flow*. See `Flow rules management`_.
2547
2548ethertype_filter
2549~~~~~~~~~~~~~~~~~~~~
2550
2551Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2552
2553   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2554                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2555
2556The available information parameters are:
2557
2558* ``port_id``: The port which the Ethertype filter assigned on.
2559
2560* ``mac_addr``: Compare destination mac address.
2561
2562* ``mac_ignr``: Ignore destination mac address match.
2563
2564* ``mac_address``: Destination mac address to match.
2565
2566* ``ether_type``: The EtherType value want to match,
2567  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2568
2569* ``queue_id``: The receive queue associated with this EtherType filter.
2570  It is meaningless when deleting or dropping.
2571
2572Example, to add/remove an ethertype filter rule::
2573
2574   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2575                             ethertype 0x0806 fwd queue 3
2576
2577   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2578                             ethertype 0x0806 fwd queue 3
2579
25802tuple_filter
2581~~~~~~~~~~~~~~~~~
2582
2583Add or delete a 2-tuple filter,
2584which identifies packets by specific protocol and destination TCP/UDP port
2585and forwards packets into one of the receive queues::
2586
2587   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2588                 protocol (protocol_value) mask (mask_value) \
2589                 tcp_flags (tcp_flags_value) priority (prio_value) \
2590                 queue (queue_id)
2591
2592The available information parameters are:
2593
2594* ``port_id``: The port which the 2-tuple filter assigned on.
2595
2596* ``dst_port_value``: Destination port in L4.
2597
2598* ``protocol_value``: IP L4 protocol.
2599
2600* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2601
2602* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2603
2604* ``prio_value``: Priority of this filter.
2605
2606* ``queue_id``: The receive queue associated with this 2-tuple filter.
2607
2608Example, to add/remove an 2tuple filter rule::
2609
2610   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2611                          tcp_flags 0x02 priority 3 queue 3
2612
2613   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2614                          tcp_flags 0x02 priority 3 queue 3
2615
26165tuple_filter
2617~~~~~~~~~~~~~~~~~
2618
2619Add or delete a 5-tuple filter,
2620which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2621and routes packets into one of the receive queues::
2622
2623   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2624                 (src_address) dst_port (dst_port_value) \
2625                 src_port (src_port_value) protocol (protocol_value) \
2626                 mask (mask_value) tcp_flags (tcp_flags_value) \
2627                 priority (prio_value) queue (queue_id)
2628
2629The available information parameters are:
2630
2631* ``port_id``: The port which the 5-tuple filter assigned on.
2632
2633* ``dst_address``: Destination IP address.
2634
2635* ``src_address``: Source IP address.
2636
2637* ``dst_port_value``: TCP/UDP destination port.
2638
2639* ``src_port_value``: TCP/UDP source port.
2640
2641* ``protocol_value``: L4 protocol.
2642
2643* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2644
2645* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2646
2647* ``prio_value``: The priority of this filter.
2648
2649* ``queue_id``: The receive queue associated with this 5-tuple filter.
2650
2651Example, to add/remove an 5tuple filter rule::
2652
2653   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2654            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2655            flags 0x0 priority 3 queue 3
2656
2657   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2658            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2659            flags 0x0 priority 3 queue 3
2660
2661syn_filter
2662~~~~~~~~~~
2663
2664Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2665
2666   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2667
2668The available information parameters are:
2669
2670* ``port_id``: The port which the SYN filter assigned on.
2671
2672* ``high``: This SYN filter has higher priority than other filters.
2673
2674* ``low``: This SYN filter has lower priority than other filters.
2675
2676* ``queue_id``: The receive queue associated with this SYN filter
2677
2678Example::
2679
2680   testpmd> syn_filter 0 add priority high queue 3
2681
2682flex_filter
2683~~~~~~~~~~~
2684
2685With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2686and routed into one of the receive queues::
2687
2688   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2689               mask (mask_value) priority (prio_value) queue (queue_id)
2690
2691The available information parameters are:
2692
2693* ``port_id``: The port which the Flex filter is assigned on.
2694
2695* ``len_value``: Filter length in bytes, no greater than 128.
2696
2697* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2698
2699* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2700
2701* ``prio_value``: The priority of this filter.
2702
2703* ``queue_id``: The receive queue associated with this Flex filter.
2704
2705Example::
2706
2707   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2708                          mask 000C priority 3 queue 3
2709
2710   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2711                          mask 000C priority 3 queue 3
2712
2713
2714.. _testpmd_flow_director:
2715
2716flow_director_filter
2717~~~~~~~~~~~~~~~~~~~~
2718
2719The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2720
2721Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2722Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2723
2724* Perfect match filters.
2725  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2726  The masked fields are for IP flow.
2727
2728* Signature filters.
2729  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2730
2731* Perfect-mac-vlan match filters.
2732  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2733  The masked fields are for MAC VLAN flow.
2734
2735* Perfect-tunnel match filters.
2736  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2737  The masked fields are for tunnel flow.
2738
2739* Perfect-raw-flow-type match filters.
2740  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2741  The masked fields are specified by input sets.
2742
2743The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2744per flow type and the flexible payload.
2745
2746The Flow Director can also mask out parts of all of these fields so that filters
2747are only applied to certain fields or parts of the fields.
2748
2749Note that for raw flow type mode the source and destination fields in the
2750raw packet buffer need to be presented in a reversed order with respect
2751to the expected received packets.
2752For example: IP source and destination addresses or TCP/UDP/SCTP
2753source and destination ports
2754
2755Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2756
2757# Commands to add flow director filters of different flow types::
2758
2759   flow_director_filter (port_id) mode IP (add|del|update) \
2760                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2761                        src (src_ip_address) dst (dst_ip_address) \
2762                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2763                        vlan (vlan_value) flexbytes (flexbytes_value) \
2764                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2765                        fd_id (fd_id_value)
2766
2767   flow_director_filter (port_id) mode IP (add|del|update) \
2768                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2769                        src (src_ip_address) (src_port) \
2770                        dst (dst_ip_address) (dst_port) \
2771                        tos (tos_value) ttl (ttl_value) \
2772                        vlan (vlan_value) flexbytes (flexbytes_value) \
2773                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2774                        fd_id (fd_id_value)
2775
2776   flow_director_filter (port_id) mode IP (add|del|update) \
2777                        flow (ipv4-sctp|ipv6-sctp) \
2778                        src (src_ip_address) (src_port) \
2779                        dst (dst_ip_address) (dst_port) \
2780                        tos (tos_value) ttl (ttl_value) \
2781                        tag (verification_tag) vlan (vlan_value) \
2782                        flexbytes (flexbytes_value) (drop|fwd) \
2783                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2784
2785   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2786                        ether (ethertype) flexbytes (flexbytes_value) \
2787                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2788                        fd_id (fd_id_value)
2789
2790   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2791                        mac (mac_address) vlan (vlan_value) \
2792                        flexbytes (flexbytes_value) (drop|fwd) \
2793                        queue (queue_id) fd_id (fd_id_value)
2794
2795   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2796                        mac (mac_address) vlan (vlan_value) \
2797                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2798                        flexbytes (flexbytes_value) (drop|fwd) \
2799                        queue (queue_id) fd_id (fd_id_value)
2800
2801   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2802                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2803                        packet (packet file name)
2804
2805For example, to add an ipv4-udp flow type filter::
2806
2807   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2808            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2809            fwd pf queue 1 fd_id 1
2810
2811For example, add an ipv4-other flow type filter::
2812
2813   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2814             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2815             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2816
2817flush_flow_director
2818~~~~~~~~~~~~~~~~~~~
2819
2820Flush all flow director filters on a device::
2821
2822   testpmd> flush_flow_director (port_id)
2823
2824Example, to flush all flow director filter on port 0::
2825
2826   testpmd> flush_flow_director 0
2827
2828flow_director_mask
2829~~~~~~~~~~~~~~~~~~
2830
2831Set flow director's input masks::
2832
2833   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2834                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2835                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2836
2837   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2838
2839   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2840                      mac (mac_value) tunnel-type (tunnel_type_value) \
2841                      tunnel-id (tunnel_id_value)
2842
2843Example, to set flow director mask on port 0::
2844
2845   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2846            src_mask 255.255.255.255 \
2847                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2848            dst_mask 255.255.255.255 \
2849                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2850
2851flow_director_flex_mask
2852~~~~~~~~~~~~~~~~~~~~~~~
2853
2854set masks of flow director's flexible payload based on certain flow type::
2855
2856   testpmd> flow_director_flex_mask (port_id) \
2857            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2858                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2859                  l2_payload|all) (mask)
2860
2861Example, to set flow director's flex mask for all flow type on port 0::
2862
2863   testpmd> flow_director_flex_mask 0 flow all \
2864            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2865
2866
2867flow_director_flex_payload
2868~~~~~~~~~~~~~~~~~~~~~~~~~~
2869
2870Configure flexible payload selection::
2871
2872   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2873
2874For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2875
2876   testpmd> flow_director_flex_payload 0 l4 \
2877            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2878
2879get_sym_hash_ena_per_port
2880~~~~~~~~~~~~~~~~~~~~~~~~~
2881
2882Get symmetric hash enable configuration per port::
2883
2884   get_sym_hash_ena_per_port (port_id)
2885
2886For example, to get symmetric hash enable configuration of port 1::
2887
2888   testpmd> get_sym_hash_ena_per_port 1
2889
2890set_sym_hash_ena_per_port
2891~~~~~~~~~~~~~~~~~~~~~~~~~
2892
2893Set symmetric hash enable configuration per port to enable or disable::
2894
2895   set_sym_hash_ena_per_port (port_id) (enable|disable)
2896
2897For example, to set symmetric hash enable configuration of port 1 to enable::
2898
2899   testpmd> set_sym_hash_ena_per_port 1 enable
2900
2901get_hash_global_config
2902~~~~~~~~~~~~~~~~~~~~~~
2903
2904Get the global configurations of hash filters::
2905
2906   get_hash_global_config (port_id)
2907
2908For example, to get the global configurations of hash filters of port 1::
2909
2910   testpmd> get_hash_global_config 1
2911
2912set_hash_global_config
2913~~~~~~~~~~~~~~~~~~~~~~
2914
2915Set the global configurations of hash filters::
2916
2917   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2918   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2919   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2920   (enable|disable)
2921
2922For example, to enable simple_xor for flow type of ipv6 on port 2::
2923
2924   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2925
2926set_hash_input_set
2927~~~~~~~~~~~~~~~~~~
2928
2929Set the input set for hash::
2930
2931   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2932   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2933   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2934   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2935   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2936   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2937   fld-8th|none) (select|add)
2938
2939For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2940
2941   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2942
2943set_fdir_input_set
2944~~~~~~~~~~~~~~~~~~
2945
2946The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2947on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2948
2949Set the input set for flow director::
2950
2951   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2952   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2953   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2954   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2955   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2956   sctp-dst-port|sctp-veri-tag|none) (select|add)
2957
2958For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2959
2960   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2961
2962global_config
2963~~~~~~~~~~~~~
2964
2965Set different GRE key length for input set::
2966
2967   global_config (port_id) gre-key-len (number in bytes)
2968
2969For example to set GRE key length for input set to 4 bytes on port 0::
2970
2971   testpmd> global_config 0 gre-key-len 4
2972
2973
2974.. _testpmd_rte_flow:
2975
2976Flow rules management
2977---------------------
2978
2979Control of the generic flow API (*rte_flow*) is fully exposed through the
2980``flow`` command (validation, creation, destruction, queries and operation
2981modes).
2982
2983Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2984features simultaneously may cause undefined side-effects and is therefore
2985not recommended.
2986
2987``flow`` syntax
2988~~~~~~~~~~~~~~~
2989
2990Because the ``flow`` command uses dynamic tokens to handle the large number
2991of possible flow rules combinations, its behavior differs slightly from
2992other commands, in particular:
2993
2994- Pressing *?* or the *<tab>* key displays contextual help for the current
2995  token, not that of the entire command.
2996
2997- Optional and repeated parameters are supported (provided they are listed
2998  in the contextual help).
2999
3000The first parameter stands for the operation mode. Possible operations and
3001their general syntax are described below. They are covered in detail in the
3002following sections.
3003
3004- Check whether a flow rule can be created::
3005
3006   flow validate {port_id}
3007       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3008       pattern {item} [/ {item} [...]] / end
3009       actions {action} [/ {action} [...]] / end
3010
3011- Create a flow rule::
3012
3013   flow create {port_id}
3014       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3015       pattern {item} [/ {item} [...]] / end
3016       actions {action} [/ {action} [...]] / end
3017
3018- Destroy specific flow rules::
3019
3020   flow destroy {port_id} rule {rule_id} [...]
3021
3022- Destroy all flow rules::
3023
3024   flow flush {port_id}
3025
3026- Query an existing flow rule::
3027
3028   flow query {port_id} {rule_id} {action}
3029
3030- List existing flow rules sorted by priority, filtered by group
3031  identifiers::
3032
3033   flow list {port_id} [group {group_id}] [...]
3034
3035- Restrict ingress traffic to the defined flow rules::
3036
3037   flow isolate {port_id} {boolean}
3038
3039Validating flow rules
3040~~~~~~~~~~~~~~~~~~~~~
3041
3042``flow validate`` reports whether a flow rule would be accepted by the
3043underlying device in its current state but stops short of creating it. It is
3044bound to ``rte_flow_validate()``::
3045
3046   flow validate {port_id}
3047      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3048      pattern {item} [/ {item} [...]] / end
3049      actions {action} [/ {action} [...]] / end
3050
3051If successful, it will show::
3052
3053   Flow rule validated
3054
3055Otherwise it will show an error message of the form::
3056
3057   Caught error type [...] ([...]): [...]
3058
3059This command uses the same parameters as ``flow create``, their format is
3060described in `Creating flow rules`_.
3061
3062Check whether redirecting any Ethernet packet received on port 0 to RX queue
3063index 6 is supported::
3064
3065   testpmd> flow validate 0 ingress pattern eth / end
3066      actions queue index 6 / end
3067   Flow rule validated
3068   testpmd>
3069
3070Port 0 does not support TCPv6 rules::
3071
3072   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3073      actions drop / end
3074   Caught error type 9 (specific pattern item): Invalid argument
3075   testpmd>
3076
3077Creating flow rules
3078~~~~~~~~~~~~~~~~~~~
3079
3080``flow create`` validates and creates the specified flow rule. It is bound
3081to ``rte_flow_create()``::
3082
3083   flow create {port_id}
3084      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3085      pattern {item} [/ {item} [...]] / end
3086      actions {action} [/ {action} [...]] / end
3087
3088If successful, it will return a flow rule ID usable with other commands::
3089
3090   Flow rule #[...] created
3091
3092Otherwise it will show an error message of the form::
3093
3094   Caught error type [...] ([...]): [...]
3095
3096Parameters describe in the following order:
3097
3098- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3099- A matching pattern, starting with the *pattern* token and terminated by an
3100  *end* pattern item.
3101- Actions, starting with the *actions* token and terminated by an *end*
3102  action.
3103
3104These translate directly to *rte_flow* objects provided as-is to the
3105underlying functions.
3106
3107The shortest valid definition only comprises mandatory tokens::
3108
3109   testpmd> flow create 0 pattern end actions end
3110
3111Note that PMDs may refuse rules that essentially do nothing such as this
3112one.
3113
3114**All unspecified object values are automatically initialized to 0.**
3115
3116Attributes
3117^^^^^^^^^^
3118
3119These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3120specified before the ``pattern`` token.
3121
3122- ``group {group id}``: priority group.
3123- ``priority {level}``: priority level within group.
3124- ``ingress``: rule applies to ingress traffic.
3125- ``egress``: rule applies to egress traffic.
3126- ``transfer``: apply rule directly to endpoints found in pattern.
3127
3128Each instance of an attribute specified several times overrides the previous
3129value as shown below (group 4 is used)::
3130
3131   testpmd> flow create 0 group 42 group 24 group 4 [...]
3132
3133Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3134
3135While not specifying a direction is an error, some rules may allow both
3136simultaneously.
3137
3138Most rules affect RX therefore contain the ``ingress`` token::
3139
3140   testpmd> flow create 0 ingress pattern [...]
3141
3142Matching pattern
3143^^^^^^^^^^^^^^^^
3144
3145A matching pattern starts after the ``pattern`` token. It is made of pattern
3146items and is terminated by a mandatory ``end`` item.
3147
3148Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3149rte_flow_item_type``).
3150
3151The ``/`` token is used as a separator between pattern items as shown
3152below::
3153
3154   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3155
3156Note that protocol items like these must be stacked from lowest to highest
3157layer to make sense. For instance, the following rule is either invalid or
3158unlikely to match any packet::
3159
3160   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3161
3162More information on these restrictions can be found in the *rte_flow*
3163documentation.
3164
3165Several items support additional specification structures, for example
3166``ipv4`` allows specifying source and destination addresses as follows::
3167
3168   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3169      dst is 10.2.0.0 / end [...]
3170
3171This rule matches all IPv4 traffic with the specified properties.
3172
3173In this example, ``src`` and ``dst`` are field names of the underlying
3174``struct rte_flow_item_ipv4`` object. All item properties can be specified
3175in a similar fashion.
3176
3177The ``is`` token means that the subsequent value must be matched exactly,
3178and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3179accordingly. Possible assignment tokens are:
3180
3181- ``is``: match value perfectly (with full bit-mask).
3182- ``spec``: match value according to configured bit-mask.
3183- ``last``: specify upper bound to establish a range.
3184- ``mask``: specify bit-mask with relevant bits set to one.
3185- ``prefix``: generate bit-mask from a prefix length.
3186
3187These yield identical results::
3188
3189   ipv4 src is 10.1.1.1
3190
3191::
3192
3193   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3194
3195::
3196
3197   ipv4 src spec 10.1.1.1 src prefix 32
3198
3199::
3200
3201   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3202
3203::
3204
3205   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3206
3207Inclusive ranges can be defined with ``last``::
3208
3209   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3210
3211Note that ``mask`` affects both ``spec`` and ``last``::
3212
3213   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3214      # matches 10.1.0.0 to 10.2.255.255
3215
3216Properties can be modified multiple times::
3217
3218   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3219
3220::
3221
3222   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3223
3224Pattern items
3225^^^^^^^^^^^^^
3226
3227This section lists supported pattern items and their attributes, if any.
3228
3229- ``end``: end list of pattern items.
3230
3231- ``void``: no-op pattern item.
3232
3233- ``invert``: perform actions when pattern does not match.
3234
3235- ``any``: match any protocol for the current layer.
3236
3237  - ``num {unsigned}``: number of layers covered.
3238
3239- ``pf``: match traffic from/to the physical function.
3240
3241- ``vf``: match traffic from/to a virtual function ID.
3242
3243  - ``id {unsigned}``: VF ID.
3244
3245- ``phy_port``: match traffic from/to a specific physical port.
3246
3247  - ``index {unsigned}``: physical port index.
3248
3249- ``port_id``: match traffic from/to a given DPDK port ID.
3250
3251  - ``id {unsigned}``: DPDK port ID.
3252
3253- ``mark``: match value set in previously matched flow rule using the mark action.
3254
3255  - ``id {unsigned}``: arbitrary integer value.
3256
3257- ``raw``: match an arbitrary byte string.
3258
3259  - ``relative {boolean}``: look for pattern after the previous item.
3260  - ``search {boolean}``: search pattern from offset (see also limit).
3261  - ``offset {integer}``: absolute or relative offset for pattern.
3262  - ``limit {unsigned}``: search area limit for start of pattern.
3263  - ``pattern {string}``: byte string to look for.
3264
3265- ``eth``: match Ethernet header.
3266
3267  - ``dst {MAC-48}``: destination MAC.
3268  - ``src {MAC-48}``: source MAC.
3269  - ``type {unsigned}``: EtherType or TPID.
3270
3271- ``vlan``: match 802.1Q/ad VLAN tag.
3272
3273  - ``tci {unsigned}``: tag control information.
3274  - ``pcp {unsigned}``: priority code point.
3275  - ``dei {unsigned}``: drop eligible indicator.
3276  - ``vid {unsigned}``: VLAN identifier.
3277  - ``inner_type {unsigned}``: inner EtherType or TPID.
3278
3279- ``ipv4``: match IPv4 header.
3280
3281  - ``tos {unsigned}``: type of service.
3282  - ``ttl {unsigned}``: time to live.
3283  - ``proto {unsigned}``: next protocol ID.
3284  - ``src {ipv4 address}``: source address.
3285  - ``dst {ipv4 address}``: destination address.
3286
3287- ``ipv6``: match IPv6 header.
3288
3289  - ``tc {unsigned}``: traffic class.
3290  - ``flow {unsigned}``: flow label.
3291  - ``proto {unsigned}``: protocol (next header).
3292  - ``hop {unsigned}``: hop limit.
3293  - ``src {ipv6 address}``: source address.
3294  - ``dst {ipv6 address}``: destination address.
3295
3296- ``icmp``: match ICMP header.
3297
3298  - ``type {unsigned}``: ICMP packet type.
3299  - ``code {unsigned}``: ICMP packet code.
3300
3301- ``udp``: match UDP header.
3302
3303  - ``src {unsigned}``: UDP source port.
3304  - ``dst {unsigned}``: UDP destination port.
3305
3306- ``tcp``: match TCP header.
3307
3308  - ``src {unsigned}``: TCP source port.
3309  - ``dst {unsigned}``: TCP destination port.
3310
3311- ``sctp``: match SCTP header.
3312
3313  - ``src {unsigned}``: SCTP source port.
3314  - ``dst {unsigned}``: SCTP destination port.
3315  - ``tag {unsigned}``: validation tag.
3316  - ``cksum {unsigned}``: checksum.
3317
3318- ``vxlan``: match VXLAN header.
3319
3320  - ``vni {unsigned}``: VXLAN identifier.
3321
3322- ``e_tag``: match IEEE 802.1BR E-Tag header.
3323
3324  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3325
3326- ``nvgre``: match NVGRE header.
3327
3328  - ``tni {unsigned}``: virtual subnet ID.
3329
3330- ``mpls``: match MPLS header.
3331
3332  - ``label {unsigned}``: MPLS label.
3333
3334- ``gre``: match GRE header.
3335
3336  - ``protocol {unsigned}``: protocol type.
3337
3338- ``fuzzy``: fuzzy pattern match, expect faster than default.
3339
3340  - ``thresh {unsigned}``: accuracy threshold.
3341
3342- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3343
3344  - ``teid {unsigned}``: tunnel endpoint identifier.
3345
3346- ``geneve``: match GENEVE header.
3347
3348  - ``vni {unsigned}``: virtual network identifier.
3349  - ``protocol {unsigned}``: protocol type.
3350
3351- ``vxlan-gpe``: match VXLAN-GPE header.
3352
3353  - ``vni {unsigned}``: VXLAN-GPE identifier.
3354
3355- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3356
3357  - ``sha {MAC-48}``: sender hardware address.
3358  - ``spa {ipv4 address}``: sender IPv4 address.
3359  - ``tha {MAC-48}``: target hardware address.
3360  - ``tpa {ipv4 address}``: target IPv4 address.
3361
3362- ``ipv6_ext``: match presence of any IPv6 extension header.
3363
3364  - ``next_hdr {unsigned}``: next header.
3365
3366- ``icmp6``: match any ICMPv6 header.
3367
3368  - ``type {unsigned}``: ICMPv6 type.
3369  - ``code {unsigned}``: ICMPv6 code.
3370
3371- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3372
3373  - ``target_addr {ipv6 address}``: target address.
3374
3375- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3376
3377  - ``target_addr {ipv6 address}``: target address.
3378
3379- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3380
3381  - ``type {unsigned}``: ND option type.
3382
3383- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3384  link-layer address option.
3385
3386  - ``sla {MAC-48}``: source Ethernet LLA.
3387
3388- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3389  link-layer address option.
3390
3391  - ``tla {MAC-48}``: target Ethernet LLA.
3392
3393Actions list
3394^^^^^^^^^^^^
3395
3396A list of actions starts after the ``actions`` token in the same fashion as
3397`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3398terminated by a mandatory ``end`` action.
3399
3400Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3401rte_flow_action_type``).
3402
3403Dropping all incoming UDPv4 packets can be expressed as follows::
3404
3405   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3406      actions drop / end
3407
3408Several actions have configurable properties which must be specified when
3409there is no valid default value. For example, ``queue`` requires a target
3410queue index.
3411
3412This rule redirects incoming UDPv4 traffic to queue index 6::
3413
3414   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3415      actions queue index 6 / end
3416
3417While this one could be rejected by PMDs (unspecified queue index)::
3418
3419   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3420      actions queue / end
3421
3422As defined by *rte_flow*, the list is not ordered, all actions of a given
3423rule are performed simultaneously. These are equivalent::
3424
3425   queue index 6 / void / mark id 42 / end
3426
3427::
3428
3429   void / mark id 42 / queue index 6 / end
3430
3431All actions in a list should have different types, otherwise only the last
3432action of a given type is taken into account::
3433
3434   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3435
3436::
3437
3438   drop / drop / drop / end # drop is performed only once
3439
3440::
3441
3442   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3443
3444Considering they are performed simultaneously, opposite and overlapping
3445actions can sometimes be combined when the end result is unambiguous::
3446
3447   drop / queue index 6 / end # drop has no effect
3448
3449::
3450
3451   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3452
3453::
3454
3455   drop / passthru / end # drop has no effect
3456
3457Note that PMDs may still refuse such combinations.
3458
3459Actions
3460^^^^^^^
3461
3462This section lists supported actions and their attributes, if any.
3463
3464- ``end``: end list of actions.
3465
3466- ``void``: no-op action.
3467
3468- ``passthru``: let subsequent rule process matched packets.
3469
3470- ``jump``: redirect traffic to group on device.
3471
3472  - ``group {unsigned}``: group to redirect to.
3473
3474- ``mark``: attach 32 bit value to packets.
3475
3476  - ``id {unsigned}``: 32 bit value to return with packets.
3477
3478- ``flag``: flag packets.
3479
3480- ``queue``: assign packets to a given queue index.
3481
3482  - ``index {unsigned}``: queue index to use.
3483
3484- ``drop``: drop packets (note: passthru has priority).
3485
3486- ``count``: enable counters for this rule.
3487
3488- ``rss``: spread packets among several queues.
3489
3490  - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3491    the same as `set_hash_global_config`_.
3492
3493  - ``level {unsigned}``: encapsulation level for ``types``.
3494
3495  - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3496    tokens are the same as `set_hash_input_set`_, except that an empty list
3497    does not disable RSS but instead requests unspecified "best-effort"
3498    settings.
3499
3500  - ``key {string}``: RSS hash key, overrides ``key_len``.
3501
3502  - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3503    conjunction with ``key`` to pad or truncate it.
3504
3505  - ``queues [{unsigned} [...]] end``: queue indices to use.
3506
3507- ``pf``: direct traffic to physical function.
3508
3509- ``vf``: direct traffic to a virtual function ID.
3510
3511  - ``original {boolean}``: use original VF ID if possible.
3512  - ``id {unsigned}``: VF ID.
3513
3514- ``phy_port``: direct packets to physical port index.
3515
3516  - ``original {boolean}``: use original port index if possible.
3517  - ``index {unsigned}``: physical port index.
3518
3519- ``port_id``: direct matching traffic to a given DPDK port ID.
3520
3521  - ``original {boolean}``: use original DPDK port ID if possible.
3522  - ``id {unsigned}``: DPDK port ID.
3523
3524- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3525
3526  - ``mpls_ttl``: MPLS TTL.
3527
3528- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3529
3530- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3531
3532  - ``nw_ttl``: IP TTL.
3533
3534- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3535
3536- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3537
3538- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3539
3540- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3541
3542- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3543
3544  - ``ethertype``: Ethertype.
3545
3546- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3547
3548  - ``vlan_vid``: VLAN id.
3549
3550- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3551
3552  - ``vlan_pcp``: VLAN priority.
3553
3554- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3555
3556  - ``ethertype``: Ethertype.
3557
3558- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3559
3560  - ``ethertype``: Ethertype.
3561
3562Destroying flow rules
3563~~~~~~~~~~~~~~~~~~~~~
3564
3565``flow destroy`` destroys one or more rules from their rule ID (as returned
3566by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3567times as necessary::
3568
3569   flow destroy {port_id} rule {rule_id} [...]
3570
3571If successful, it will show::
3572
3573   Flow rule #[...] destroyed
3574
3575It does not report anything for rule IDs that do not exist. The usual error
3576message is shown when a rule cannot be destroyed::
3577
3578   Caught error type [...] ([...]): [...]
3579
3580``flow flush`` destroys all rules on a device and does not take extra
3581arguments. It is bound to ``rte_flow_flush()``::
3582
3583   flow flush {port_id}
3584
3585Any errors are reported as above.
3586
3587Creating several rules and destroying them::
3588
3589   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3590      actions queue index 2 / end
3591   Flow rule #0 created
3592   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3593      actions queue index 3 / end
3594   Flow rule #1 created
3595   testpmd> flow destroy 0 rule 0 rule 1
3596   Flow rule #1 destroyed
3597   Flow rule #0 destroyed
3598   testpmd>
3599
3600The same result can be achieved using ``flow flush``::
3601
3602   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3603      actions queue index 2 / end
3604   Flow rule #0 created
3605   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3606      actions queue index 3 / end
3607   Flow rule #1 created
3608   testpmd> flow flush 0
3609   testpmd>
3610
3611Non-existent rule IDs are ignored::
3612
3613   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3614      actions queue index 2 / end
3615   Flow rule #0 created
3616   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3617      actions queue index 3 / end
3618   Flow rule #1 created
3619   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3620   testpmd>
3621   testpmd> flow destroy 0 rule 0
3622   Flow rule #0 destroyed
3623   testpmd>
3624
3625Querying flow rules
3626~~~~~~~~~~~~~~~~~~~
3627
3628``flow query`` queries a specific action of a flow rule having that
3629ability. Such actions collect information that can be reported using this
3630command. It is bound to ``rte_flow_query()``::
3631
3632   flow query {port_id} {rule_id} {action}
3633
3634If successful, it will display either the retrieved data for known actions
3635or the following message::
3636
3637   Cannot display result for action type [...] ([...])
3638
3639Otherwise, it will complain either that the rule does not exist or that some
3640error occurred::
3641
3642   Flow rule #[...] not found
3643
3644::
3645
3646   Caught error type [...] ([...]): [...]
3647
3648Currently only the ``count`` action is supported. This action reports the
3649number of packets that hit the flow rule and the total number of bytes. Its
3650output has the following format::
3651
3652   count:
3653    hits_set: [...] # whether "hits" contains a valid value
3654    bytes_set: [...] # whether "bytes" contains a valid value
3655    hits: [...] # number of packets
3656    bytes: [...] # number of bytes
3657
3658Querying counters for TCPv6 packets redirected to queue 6::
3659
3660   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3661      actions queue index 6 / count / end
3662   Flow rule #4 created
3663   testpmd> flow query 0 4 count
3664   count:
3665    hits_set: 1
3666    bytes_set: 0
3667    hits: 386446
3668    bytes: 0
3669   testpmd>
3670
3671Listing flow rules
3672~~~~~~~~~~~~~~~~~~
3673
3674``flow list`` lists existing flow rules sorted by priority and optionally
3675filtered by group identifiers::
3676
3677   flow list {port_id} [group {group_id}] [...]
3678
3679This command only fails with the following message if the device does not
3680exist::
3681
3682   Invalid port [...]
3683
3684Output consists of a header line followed by a short description of each
3685flow rule, one per line. There is no output at all when no flow rules are
3686configured on the device::
3687
3688   ID      Group   Prio    Attr    Rule
3689   [...]   [...]   [...]   [...]   [...]
3690
3691``Attr`` column flags:
3692
3693- ``i`` for ``ingress``.
3694- ``e`` for ``egress``.
3695
3696Creating several flow rules and listing them::
3697
3698   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3699      actions queue index 6 / end
3700   Flow rule #0 created
3701   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3702      actions queue index 2 / end
3703   Flow rule #1 created
3704   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3705      actions rss queues 6 7 8 end / end
3706   Flow rule #2 created
3707   testpmd> flow list 0
3708   ID      Group   Prio    Attr    Rule
3709   0       0       0       i-      ETH IPV4 => QUEUE
3710   1       0       0       i-      ETH IPV6 => QUEUE
3711   2       0       5       i-      ETH IPV4 UDP => RSS
3712   testpmd>
3713
3714Rules are sorted by priority (i.e. group ID first, then priority level)::
3715
3716   testpmd> flow list 1
3717   ID      Group   Prio    Attr    Rule
3718   0       0       0       i-      ETH => COUNT
3719   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3720   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3721   1       24      0       i-      ETH IPV4 UDP => QUEUE
3722   4       24      10      i-      ETH IPV4 TCP => DROP
3723   3       24      20      i-      ETH IPV4 => DROP
3724   2       24      42      i-      ETH IPV4 UDP => QUEUE
3725   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3726   testpmd>
3727
3728Output can be limited to specific groups::
3729
3730   testpmd> flow list 1 group 0 group 63
3731   ID      Group   Prio    Attr    Rule
3732   0       0       0       i-      ETH => COUNT
3733   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3734   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3735   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3736   testpmd>
3737
3738Toggling isolated mode
3739~~~~~~~~~~~~~~~~~~~~~~
3740
3741``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3742must only be injected from the defined flow rules; that no default traffic
3743is expected outside those rules and the driver is free to assign more
3744resources to handle them. It is bound to ``rte_flow_isolate()``::
3745
3746 flow isolate {port_id} {boolean}
3747
3748If successful, enabling or disabling isolated mode shows either::
3749
3750 Ingress traffic on port [...]
3751    is now restricted to the defined flow rules
3752
3753Or::
3754
3755 Ingress traffic on port [...]
3756    is not restricted anymore to the defined flow rules
3757
3758Otherwise, in case of error::
3759
3760   Caught error type [...] ([...]): [...]
3761
3762Mainly due to its side effects, PMDs supporting this mode may not have the
3763ability to toggle it more than once without reinitializing affected ports
3764first (e.g. by exiting testpmd).
3765
3766Enabling isolated mode::
3767
3768 testpmd> flow isolate 0 true
3769 Ingress traffic on port 0 is now restricted to the defined flow rules
3770 testpmd>
3771
3772Disabling isolated mode::
3773
3774 testpmd> flow isolate 0 false
3775 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3776 testpmd>
3777
3778Sample QinQ flow rules
3779~~~~~~~~~~~~~~~~~~~~~~
3780
3781Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3782
3783   testpmd> port stop 0
3784   testpmd> vlan set qinq on 0
3785
3786The above command sets the inner and outer TPID's to 0x8100.
3787
3788To change the TPID's the following commands should be used::
3789
3790   testpmd> vlan set outer tpid 0xa100 0
3791   testpmd> vlan set inner tpid 0x9100 0
3792   testpmd> port start 0
3793
3794Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3795
3796::
3797
3798   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3799       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3800   Flow rule #0 validated
3801
3802   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3803       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3804   Flow rule #0 created
3805
3806   testpmd> flow list 0
3807   ID      Group   Prio    Attr    Rule
3808   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3809
3810Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3811
3812::
3813
3814   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3815        vlan tci is 654 / end actions pf / queue index 0 / end
3816   Flow rule #1 validated
3817
3818   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3819        vlan tci is 654 / end actions pf / queue index 1 / end
3820   Flow rule #1 created
3821
3822   testpmd> flow list 0
3823   ID      Group   Prio    Attr    Rule
3824   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3825   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3826